DataScientist

Customer Revenue Analysis for Retail Store
 Predictive modeling with Linear Regression, Decision Trees using Python.
 Applied PCA and Regularization techniques to predict revenue with an rmse of 1.150943e+03
to increase the revenue of the retail store.

Health Condition Analysis of a person
 Built Logistic Regression, Decision Trees ,Random Forest and XGBoost using Python.
 Applied PCA and Regularization techniques, StepAIC, Checked on MultiCollinearity of the
columns.
 Achieved an accuracy of 90.42% by using the transformation techniques on the features.

T-shirts Image Classification by their graphic type
 Built CNN, RNN, and LSTM Neural Network models to predict appropriate category from 21
classes based on the summary and description.
 Achieved 97.93?curacy in CNN using Transfer Learning. Monitored loss for early stopping,
used drop out, and an optimizer to get best results

Bankruptcy Prediction of a Firm using a Machine Learning Algorithms
 Built Adaboost(95.3), GradientBoost(94.20) for achieving better accuracy.
Predicting the urgency of the ticket with respect to the body of the ticket using neural networks
 Built CNN, RNN, and LSTM Neural Network models to predict appropriate category from 2
classes based on the body of the ticket.
 Achieved an accuracy of 84% using LSTM Network.

Customer Subscription Prediction to a term deposit in a bank
 Built Logistic Regression, Decision Trees ,Random Forest and XGBoost using PySpark
 Used Plotly for EDA.
 Achieved an accuracy of 90.12% by using the transformation techniques on the features.
Predicting a customer churn using Spark MLlib (Spark’s machine learning library).
 Built Logistic Regression,Random Forest using PySpark
 Achieved an accuracy of 77.9 % by using Decision Tree.
 Used Bucketizer feature transformer available in pyspark.ml.feature.

RainFall Prediction using a binary classification model
 Built Logistic Regression,Random Forest ,NaiveBayes ,XGBoost using PySpark
 Achieved an accuracy of 96.9 % by using the transformation techniques on the features.

Analysis of Hotel Reviews, Segmentation & Developing a Recommender System:
 Performed analysis based on hotel reviews and segmentation of hotels (based on reviews and
other features that you might feel appropriate to derive)such that the user might be informed
about the similar hotels nearby.
 Used Tableau for Data Visualizations.
 Used Multinomial Naïve Bayes,Logistic Regression for building the classification system.
 Used kmeans clustering for Segmentation of Hotels.
 Recommended hotels in a specific location to the users based on their previous preferences
stated in user ratings.Used Collaborative Filtering for building this.

Image Classification :
 Built a binary image classifier that can predict to classify if a given image is a cat or a dog. The
data consists of train and test sets with a volume of 6000 images.
 CNN coupled with a Feed Forward Neural Network architecture in Python (Keras API).

Professional Experience

Project #1:

Churn Analysis: Nation Wide Building Society

Role: DataScientist

Organization: Infosys Technologies

Analytical Tools Used: R, Python

This Project mainly focuses on predicting the future default customers based on the Various data sources

• Identification of appropriate fields and collection of data from bank.
• Understanding the business requirement of analytical problem Data cleansing/Data Preprocessing.
• Deliver updates and performance tracking of solutions to business team
• Performed exploratory data analysis to identify the sample dependencies
• Apply the best predictive model and scored the production data

TeamLead

Responsibilities:

Involved in Initial designing and creating Use case diagrams, Sequence Diagrams and class diagrams using the MS Visio Tool.
Worked as a Java Developer and involved in analysis of requirements, design, development, integration, deployment and testing and fixing the production issues.

Designing, building and maintaining high-performance
Developed Java applications using spring and other J2EE design patterns.

Configured Coherence Cache to improve the performance of the application.
Configured using JDBC for the Database connection and SQL to add and retrieve the Information from the Database
Involved in the development of Functions, Triggers for PL/SQL backend implementation.
Mentoring the team in development of Java related classes and components using Eclipse as IDE for developing the application.
Developed Unit Tests and Test Suites using Junit and used Log4j for traces and logs
Used Perforce for software configuration management and version control
Developed a wiki in the sharepoint where all the details like Server info and the functionality of the application is described.
Developed Oracle Scripts as per the data model.
Created service applications using Webservices and JMS.
Involved in developing core persistence classes using Hibernate framework, writing HQL queries, creating hibernate mapping (.hbm) files, DB schema.
Worked on the SOA, WSDL, XML, Schema, SOAP and WebServices calls to integrate the service applications.
Worked on implementing Caching by using ContinousQueryCache ,Entry Processors.

Senior Java Developer

Responsibilities:

Complete ownership of the application as the only Java developer working as part of the project
Requirements analysis and design using UML

Involved in Initial designing and creating Use case diagrams, Sequence Diagrams and class diagrams using the MS Visio Tool.
Used ClearCase for software configuration management and version control
Used Junit Test Cases for testing the output file
Used Ant scripts for building and packaging the application.
Designed and developed back-end store procedures with Oracle PL/SQL.
Designed and implemented unit testing for all developed code.
Worked with J2EE transactions for data integrity across multiple systems

ModuleLead

Responsibilities:

Working as a Developer and involved in analysis of requirements, design and development and supporting the project (implementing enhancements and Change Requests).
Implemented MVC-Model architecture using spring and other J2EE design patterns like Service Locator, Session Façade, Singleton and Factory Pattern for the application development.
Interact with business analysts to convert business requirements to technical requirements, Followed Agile methodology and SCRUM meetings to track, optimize and tailor features to customer needs.
Involved in designing the framework using sequence diagrams, class diagrams and domain model using RSA
Used Ant build script to deploy the application and JUnit to test unit test cases
Provided Production Support for the application.
Developed web application using Struts Framework;
Developed user interfaces using JSP, HTML and CSS
Used JavaScript and struts validation framework for performing front end validations.
Used ClearCase for software configuration management and version control
Deployed the application on the Weblogic Application Server

Module Lead

Responsibilities:

Worked as a Lead Developer and involved in analysis of requirements, design, development, integration, deployment and testing and fixing the production issues.
Implemented MVC-Model architecture using spring and other J2EE design patterns for the application development.
Configured using JDBC for the Database connection and SQL to add and retrieve the Information from the Database.
Developed web application using Struts Framework.
Refractored the java code efficiently to suit the new database
Created JSP’s, Conversations, POJOs, Services ,DAO and Session beans.
Developed JSP pages using Struts tag libraries for dynamically displaying the data on the browser and developed Custom tag libraries
Used for development of Java related classes and components using Eclipse 3.5 as IDE for developing the application.
Involved in developing ANTbuildscript to automate the compilation, creating directory structures, JAR and EAR files.
Developed Unit Tests and Test Suites using Junit and used Log4j for traces and logs
Used ClearCase for software configuration management and version control
Deployed the application on the Web logic Application Server.

Senior Java Developer

Responsibilities:

Played the role of a Team Lead and took complete ownership of three separate modules in the application.
Requirement gathering Analysis and Design of the application.
Front end UI design and prototypes using HTML and JavaScript.
Developed front-end screens using JSP and JavaScript technologies.
Developed the application using Spring Core framework.
Database design and data model for the web application.
Experience in developing database objects and programs such as stored procedures and packages using PL/SQL
Experience in application deployment on Web sphere application server.
Also played the role of Quality Analyst in testing the applications.
Developed Unit Tests and Test Suites using JUnit for testing the developed Module.
VSS used for version control.
Technical Design, documentation, review and rework
Working with team, and providing coding standards and inputs on delivery standards.
Working towards improving quality and productivity by planning and scheduling tasks appropriately.
Delivery responsibility of all Project requests.

Client: PRNewswire

Role: Java Developer

Technologies: Java 1.5, Liferay, JSP,Log4j,MYSQL,Eclipse

Responsibilities:

Involved in Design and documentation.
Development of end to end functionality using Liferay framework.
Development of Photo uploading metadata functionality of the application.
Followed agile methodology. Implemented the SCRUM process.
Used VSS for version control and Configuration Management.

Role:Java Developer

Responsibilities:

Analysis of customer requirements based on the original draft.
Worked as a developer for the following SQP’s: INAV R1,R2,R3,R4,R5 ,HK –Venus, AU-Pay Flex ,Mexico Transparency, Germany SBS Platinum ,JADR .
Performed Unit Testing and Supported during various phases of the project till Launch.
Extended my support to other SQP’s for Testing and Quality work. Delivered all artifacts or deliverables with zero logic defects successfully within timelines.
Involved in requirement gathering, analysis and development of migrating the Logging mechanism of the whole project.
Used CVS for Configuration Management. Received appreciation mails for delivering the project.

Predictive modeling with Linear Regression, Decision Trees using Python. Applied PCA and Regularization techniques to predict revenue with an rmse of 1.150943e+03 to increase the revenue of the retail store.

Built Logistic Regression, Decision Trees, Random Forest and XGBoost using Python. Applied PCA and Regularization techniques, StepAIC, Checked on MultiCollinearity of the columns. Achieved an accuracy of 90.42 by using the transformation techniques on the features.

Built CNN, RNN, and LSTM Neural Network models to predict appropriate category from 21 classes based on the summary and description. Achieved 97.93 accuracy in CNN using Transfer Learning. Monitored loss for early stopping, used drop out, and an optimizer to get best results.

Built Adaboost(95.3), GradientBoost(94.20) for achieving better accuracy.

Built CNN, RNN, and LSTM Neural Network models to predict appropriate category from 2 classes based on the body of the ticket. Achieved an accuracy of 84 using LSTM Network.

Built Logistic Regression, Decision Trees, Random Forest and XGBoost using PySpark. Used Plotly for EDA. Achieved an accuracy of 90.12 by using the transformation techniques on the features.

Built Logistic Regression, Random Forest using PySpark. Achieved an accuracy of 77.9 by using Decision Tree. Used Bucketizer feature transformer available in pyspark.ml.feature.

Built Logistic Regression, Random Forest, NaiveBayes, XGBoost using PySpark. Achieved an accuracy of 96.9 by using the transformation techniques on the features.

Performed analysis based on hotel reviews and segmentation of hotels such that the user might be informed about similar hotels nearby. Used Tableau for Data Visualizations. Used Multinomial Nave Bayes, Logistic Regression for building the classification system. Used kmeans clustering for Segmentation of Hotels. Recommended hotels in a specific location to the users based on their previous preferences stated in user ratings. Used Collaborative Filtering for building this.

Built a binary image classifier that can predict to classify if a given image is a cat or a dog. The data consists of train and test sets with a volume of 6000 images. CNN coupled with a Feed Forward Neural Network architecture in Python (Keras API).

Project mainly focuses on predicting the future default customers based on Various data sources. Identification of appropriate fields and collection of data from bank. Understanding the business requirement of analytical problem Data cleansing/Data Preprocessing. Deliver updates and performance tracking of solutions to business team. Performed exploratory data analysis to identify the sample dependencies. Apply the best predictive model and scored the production data. Analytical Tools Used R, Python.